The pancreas operates as two interconnected systems: one for digestion, exocrine, and one for blood sugar regulation, endocrine. While usually studied in isolation, these systems exert significant mechanical and chemical influence on one another. This project proposes a physiologically based pharmacokinetic model to simulate how digestive enzyme levels physically stress or support insulin-producing Beta cells.
This mechanistic approach provides a predictive framework for understanding how localized inflammation in pancreatitis, structural pressure from pancreatic cancer, or enzymatic shifts trigger the onset of Type 3c diabetes. By integrating chemical reaction rates with cellular data using Neural Ordinary Differential Equations, the goal is to quantify the thresholds where digestive dysfunction leads to endocrine collapse. This synthesis of computer science and molecular kinetics offers a predictive mechanistic framework for diagnosing and managing the progression of multi-system pancreatic diseases.
